Proton transfer reaction time-of-flight mass spectrometry (PTR-ToF-MS) represents an attractive tool for the real-time analysis of VOC profiles in human breath. Quantification of breath VOCs by means of direct MS may be affected by the matrix, as human breath not only contains several hundred VOCs at the ppbV-pptV level, but is water saturated and contains percentage levels of CO2. Investigation of breath biomarkers in clinical studies requires quantitative and comparable results. We therefore systematically assessed the effect of humidity, CO2 and O2 on the results of PTR-MS analysis. We investigated more than 20 VOCs, including aldehydes, ketones, aromatic compounds and hydrocarbons with different sample humidity, CO2 and O2 content. The influence of data processing (e.g. normalization to the H3O+ ion count) was also addressed. An increase of the H3O+ count of about 20% was observed when the humidity in the sample was increased to breath levels. Large differences regarding the measured VOC intensities were found between the dry and humid samples. Data normalization to the H3O+ or water-clusters could not fully compensate for the humidity-dependent effects. However, as the determination of most VOCs linearly depends on the humidity over the whole investigated range, factor-based correction seems possible. The effects of CO2 were more pronounced in the dry samples than in the humid samples but only had a minor influence on the results. The same was true for the influence of O2. For the reliable quantification of VOCs in clinical studies and for the standardization of VOC research, well-adapted calibration standards are required for PTR-MS analysis.